Stephen Finucane wrote:
On Wed, 2022-10-19 at 17:10 +0000, Elõd Illés wrote:
During 'TC + Community leaders interaction' [1] a case was discussed, where a late library release caused last minute fire fighting in Zed cycle, and people discussed the possibility to introduce a (non-client) library *feature* freeze at Milestone-2 to avoid similar issues in the future. [...]
Repeating what I said on the reviews, I'd really rather not do this. [...]
I tend to agree with Stephen on this... Cutting a significant window of time to handle a relatively rare occurrence might not be a great tradeoff. From a release management perspective, if that ensured that we'd always avoid last-minute release issues, I would support it. But reality is, this covers just a part of our blind spot. There are a lot of reasons why CI for a particular project ends up no longer working, Oslo breaking changes is just one of them. Periodically checking that CI works for all projects (ideally through normalized periodic testing, if not through regularly posting a bunch of noop test changes in inactive repositories) would detect issues earlier and cover all of our blind spot. We should also make sure we only ship actively-maintained projects, so that we know who to turn to to fix it when it's broken. Freezing Oslo a lot earlier? Not convinced. -- Thierry Carrez (ttx)